@netnodeag/kraftwerk@0.36.0
Open-source agentic workspace for teams — where people and AI agents get work done together: agents (Claude Code, Codex, Pi), skills, shared knowledge, repeatable workflows, and deterministic verification
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThe package is not proven malware because the dangerous actions are exposed as explicit framework operations rather than an automatic lifecycle payload. It should still warn because it can orchestrate agents, shell commands, credentials, and remote code-bearing projects.
Decision evidence
public snapshot- The CLI and inspector can launch workflow runs that execute user-provided agent prompts and shell scripts.
- Sandbox runs forward configured API-key environment variables and, when the SSH option is selected, mount an SSH-agent socket and known-hosts file.
- The inspector can create, modify, and delete agent definition files under the configured project agent directory.
- Remote project support clones and refreshes Git repositories in a local cache.
- The flagged bundled assets are inspectable front-end JavaScript, including Vite and SQL/VBScript editor modules.
- The manifest has no preinstall, install, or postinstall hook; its only lifecycle hook is prepublishOnly.
- There is no runtime dependency on this package itself.
- The sensitive Docker forwarding path requires an invoked sandbox run and the explicit SSH option.
- No reviewed source established credential exfiltration, a covert network endpoint, or automatic consumer-project mutation.
Source & flagged code
9 flaggedPackage source references shell execution.
inspector/dist/assets/editor-7ImnDC3D.jsView on unpkg · L114Findings
Affected versions and remediation
This report applies to @netnodeag/kraftwerk@0.36.0.
See all versions of @netnodeag/kraftwerk for other recorded verdicts.
- Review the evidence and your use of @netnodeag/kraftwerk@0.36.0 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.